License plate Character Segmentation using horizontal and vertical projection with dynamic thresholding

Intelligent Traffic Systems (ITS) is an integral component of modern day road transportation networks. Identification of vehicles is one of the most important challenges to be addressed in the design of any intelligent traffic system. License plate being a unique identity for any registered vehicle, License Plate Recognition (LPR) systems have been used as the means to resolve the issue of identification of vehicles. Intelligent systems involving LPR has been widely applied in several applications such as intelligent traffic rule enforcement, prevention of car theft, intelligent traffic emergency/accident handling, monitoring of vehicle traffic and flow control, intelligent parking, automated toll payment, intelligent surveillance and security enforcement etc. A typical LPR process consists of three stages viz; License Plate Extraction, Character Segmentation and Character Recognition. This paper address the Character Segmentation problem. The character segmentation method using horizontal and vertical projection and with dynaic thresholding is proposed in this paper. From the experimental studies made, it is learned that the proposed technique works reasonably well in real world scenarios.